Skip to content
Founder Tools

Founder storiesHardware prototyping

Why James Dyson strapped cardboard to a vacuum cleaner

A sawmill gave James Dyson a way to separate dust. The 550 machines made under an early license showed how far a working prototype was from a lasting business.

James Dyson looks through the circular door of his ContraRotator washing machine, a hand on each side.
James Dyson with his ContraRotator washing machine in 2000, seven years after the DC01 vacuum launched. Photo by Michiel Hendryckx · CC BY-SA 3.0 · Resized and converted to WebP..

Lessons for founders

  • Dyson tried an industrial separation technique on a vacuum he already owned. Cardboard was enough to see whether dust would separate at that scale.
  • An early license produced 550 machines. Dyson kept looking for a way to sell the design, first through another partner and later under his own name.
From the factory floor to the home
1978
A timber-mill cyclone inspires a factory fix
1983–84
Rotork’s licensed Cyclon reaches production
1986
Apex starts producing the G-Force in Japan
1993
The Dyson DC01 launches

The early Rotork Cyclon and the Japanese G-Force were distinct products. Both preceded the DC01.[1][2][3]

The idea arrived on a factory floor

In 1978, paint powder kept clogging a cloth at the factory where James Dyson made the Ballbarrow, his redesign of the wheelbarrow. At a nearby timber mill he had seen machinery that separated sawdust from air using a cyclone. He built a 25-foot separator for his own factory to collect the powder.

At home, a vacuum started losing suction. Dyson opened its bag and found the pores clogged with dust. The factory problem and the household problem had something unusually specific in common. Both machines needed to move air while catching small particles that could block its path. That observation gave him a borrowed mechanism to try.

Sources12

Keep the vacuum and change one part

Dyson made a small cyclone out of cardboard and strapped it to his existing vacuum in place of the bag. He recalled that the arrangement picked up more dust. The motor, handle and most of the machine were already there. The improvised part let him test whether the separation principle survived the change in scale.

A good result still left a long list of unknowns. How much dust could the chamber hold? What would happen after repeated use? Would someone buy the thing? The familiar figure of 5,127 prototypes comes from Dyson’s account of developing the vacuum. The cardboard experiment was valuable for giving him a reason to make prototype two.

Sources12

What 550 machines could tell him

Rotork licensed the design. In his autobiography, Dyson says 550 Kleeneze Rotork Cyclons were made in 1983 and 1984. He describes the project as a commercial failure. He had found someone who could make the machine. He still needed a way to sell enough of them to keep going.

Another partner, Apex, made the G-Force in Japan starting in 1986. The DC01, sold under Dyson’s own name, followed in 1993. For a founder approaching a potential licensee, the early Rotork run raises a question a prototype cannot answer. Who at the partner will take responsibility for getting the finished units into customers’ hands?

Sources123

Which part deserves your first prototype?

Imagine designing a bicycle pannier that commuters can remove with one gloved hand. One unknown could kill the idea early. Will the latch stay closed on a rough road and still release with cold fingers? Bolt a candidate latch onto a borrowed rack and a plain box. Agree on a load and a release test before ordering fabric or paying for a custom shell.

Then find the person who would have to buy, stock or fit the finished bag. Ask what evidence they would need to try a small batch. A working latch gives you something to demonstrate. Their answer tells you whether the next expense should be a better latch, a sample for a shop or a conversation with a different buyer.

Sources

3 references
  1. How we made the Dyson vacuum cleaner

    The Guardian, interview with James Dyson